Revision 45878
Added by Eri Katsari about 7 years ago
| LinkCustomReducer.java | ||
|---|---|---|
| 13 | 13 |
public class LinkCustomReducer extends Reducer<Text, Text, Text, Text> {
|
| 14 | 14 |
private Logger log = Logger.getLogger(LinkCustomReducer.class); |
| 15 | 15 |
private RedisUtils redisUtils; |
| 16 |
private static final String LINE_SEPERATOR = "\t.\t"; |
|
| 17 |
private static final String FIELD_DELIM = "\t"; |
|
| 18 |
private static final String OPTIMAL_BLOCKS = "optimalBlockSize"; |
|
| 19 |
private long optimalBlockSize; |
|
| 20 | 16 |
private static final String SEPERATOR = ","; |
| 21 | 17 |
private static double RECORD_SIMILARITY_THRESHOLD = 0.8; |
| 22 | 18 |
|
| ... | ... | |
| 34 | 30 |
|
| 35 | 31 |
try {
|
| 36 | 32 |
redisUtils = new RedisUtils(context); |
| 37 |
optimalBlockSize = Integer.valueOf(redisUtils.getValue(OPTIMAL_BLOCKS)); |
|
| 38 |
log.info("OPTIMAL BLOCK SIZE " + optimalBlockSize);
|
|
| 39 | 33 |
} catch (Exception e) {
|
| 40 |
log.error("Error computing stats" + e.toString());
|
|
| 34 |
log.error("Error connecting to Redis " + e.toString());
|
|
| 41 | 35 |
throw new RuntimeException(e); |
| 42 | 36 |
} |
| 43 | 37 |
|
| ... | ... | |
| 47 | 41 |
protected void reduce(final Text key, final Iterable<Text> values, final Context context) throws IOException, InterruptedException {
|
| 48 | 42 |
//each item in the list is a block with the given key |
| 49 | 43 |
for (Text block : values) {
|
| 50 |
if (!key.toString().matches("^\\d{4}$")) {
|
|
| 51 |
context.getCounter(LINK_RECUDER_COUNTERS.BLOCKS).increment(1); |
|
| 52 |
try {
|
|
| 53 |
linkRecords(block.toString(), context); |
|
| 54 |
} catch (Exception e) {
|
|
| 55 |
log.error("Error compators records" + e.toString());
|
|
| 56 |
throw new RuntimeException(e); |
|
| 57 |
} |
|
| 58 |
|
|
| 44 |
context.getCounter(LINK_RECUDER_COUNTERS.BLOCKS).increment(1); |
|
| 45 |
try {
|
|
| 46 |
linkRecords(block.toString(), context); |
|
| 47 |
} catch (Exception e) {
|
|
| 48 |
log.error("Error comparing records" + e.toString());
|
|
| 49 |
throw new RuntimeException(e); |
|
| 59 | 50 |
} |
| 60 | 51 |
} |
| 61 | 52 |
} |
| ... | ... | |
| 65 | 56 |
String[] split = block.split(SEPERATOR); |
| 66 | 57 |
Map<String, String> sourceRecords = new HashMap<>(); |
| 67 | 58 |
Map<String, String> targetRecords = new HashMap<>(); |
| 59 |
|
|
| 68 | 60 |
for (String recordId : split) {
|
| 69 | 61 |
if (recordId.contains("source_")) {
|
| 70 | 62 |
sourceRecords.put(recordId, ""); |
| 71 | 63 |
context.getCounter(LINK_RECUDER_COUNTERS.SOURCE_TRIPLES).increment(1); |
| 72 | 64 |
} else {
|
| 73 |
targetRecords.put(recordId, "");
|
|
| 65 |
targetRecords.put(recordId, redisUtils.getValue(recordId));
|
|
| 74 | 66 |
context.getCounter(LINK_RECUDER_COUNTERS.TARGET_TRIPLES).increment(1); |
| 75 | 67 |
} |
| 76 | 68 |
} |
| ... | ... | |
| 82 | 74 |
// |
| 83 | 75 |
|
| 84 | 76 |
//TODO |
| 85 |
//String targetRecord = redisUtils.getValue(targetId); |
|
| 86 |
|
|
| 87 |
if (targetRecords.get(targetId).isEmpty()) {
|
|
| 88 |
targetRecords.put(targetId, redisUtils.getValue(targetId)); |
|
| 89 |
} |
|
| 90 |
|
|
| 91 |
|
|
| 77 |
//String targetRecord = redisUtils.getValue(targetId);} |
|
| 92 | 78 |
String targetRecord = targetRecords.get(targetId); |
| 93 | 79 |
double similarity = MyComparator.findMatchingPair(sourceRecord, targetRecord); |
| 94 | 80 |
context.getCounter(LINK_RECUDER_COUNTERS.COMPARISONS).increment(1); |
| 95 | 81 |
if (similarity >= RECORD_SIMILARITY_THRESHOLD) {
|
| 96 | 82 |
try {
|
| 97 |
context.write(new Text(sourceId), new Text(targetId + "," + similarity));
|
|
| 83 |
context.write(new Text(sourceId), new Text(targetId + SEPERATOR + similarity));
|
|
| 98 | 84 |
context.getCounter(LINK_RECUDER_COUNTERS.WRITTEN_OUT_ENTITIES).increment(1); |
| 99 | 85 |
} catch (Exception ex) {
|
| 100 | 86 |
log.error("Error while writing records to output : " + ex.toString());
|
Also available in: Unified diff
Added M/R step for verification